We will be building a Workout Assistant which will help us maintain the right postures while doing various workout exercises. This is mainly a Pose Estimation, and Pose Matching problem. We will be using the POSENET model for Pose Estimation.
Here are some links for your reference:
Pose Detection | ML Kit | Google Developers - https://developers.google.com/ml-kit/vision/pose-detection
Pose estimation | TensorFlow Lite -https://www.tensorflow.org/lite/models/pose_estimation/overview
Pose Model - Teachable Machines - https://teachablemachine.withgoogle.com/train/pose
3D Human Pose Estimation in AI Fitness Coach Apps (mobidev.biz) - https://mobidev.biz/blog/human-pose-estimation-ai-personal-fitness-coach
Evaluation of PoseNet for applied AI-Fitness applications - https://medium.com/optima-ai/evaluation-of-posenet-for-applied-ai-in-fitness-applications-56de98d6c4e4
This will be a Flutter App leveraging the ML Models created and integrating it in the app.
Base App is uploaded.
- Fork this Repository.
- Select an issue on which you want to work. Discuss with the mentors.
- Whenever submitting a Pull Request or adding a feature first create a seperate branch using
git checkout -b <name-of-the-new-branch> - To commit changes:
git commit -m "<a meaningful message> - Before pushing your changes always make sure your local repository is updated.
git fetch origingit merge origin - If there are merge conflicts resolve them and then push changes.
- To push the changes:
git push origin <your-branch-name> - Create a Pull Request on the Repository and we will review it.